Electrowinning of metals from molten salt electrolytes involves numerous difficulties. A typical process is the production of aluminum by the Hall-Heroult process which involves the electrolysis of alumina in a molten cryolite-based bath usng carbon anodes. These carbon anodes are consumed by the anodic oxidation process with the formation of CO2/CO and their life-time is very short, typically about two to three weeks for the pre-baked type or anode. They may also add impurities to the bath. There have been numerous suggestions for non-consumable anode compositions based on various ceramic oxides and oxycompounds usually with added electro-conductive agents and electrocatalysts. Many difficulties have been encountered in practice with such anodes, the major difficulty being that the anodes are invariably consumed more or less slowly and undesirably contaminate the molten bath and the aluminum or other metal produced.
For example, U.S. Pat. Nos. 4,146,438 and 4,187,155 describe molten-salt electrolysis anodes consisting of a ceramic oxycompound matrix with an oxide or metallic conductive agent and a surface coating of an electrocatalyst e.g. oxides of cobalt, nickel, manganese, rhodium, iridium, ruthenium and silver. One of the problems with these electrodes is that the catalytic coating wears away.
Another approach, described in U.S. Pat. Nos. 3,562,135, 3,578,580 and 3,692,645, was to separate the anode and cathode by an oxygen-ion conducting diaphragm, typically made of stabilized zirconium oxide or other refractory oxides with a cubic (fluorite) lattice, including thorium oxide/uranium oxide and cerium oxide suitably stabilized with calcium oxide or magnesium oxide. In one arrangement, the ion-conductive diaphragm was applied to the operative anode surface which was either liquid or was porous, perforated or reticulated and provided with means for releasing the oxygen generated at the anode under the diaphragm. This involved considerable problems in anode design and in manufacture of the composite anode/diaphragm. Another arrangement was to separate the diaphragm from the anode surface; here, it would appear that tests failed to identify any feasible diaphragm material.